Hello,
I just received a Mcintosh MC2505 stereo 50 watts power amp.
It has a seperate winding to supply the output stage with + and - 40 volts DC.
Power consumption is 250 watts at rated output so it will be class AB or just B?
Because i did use chokes in the past for my 30 watts Hiraga amp i wanna try adding two chokes to change the present power supply from just a single cap per rail to a CLC set up.
The original cap is 10000 µF so i will probably just add another and a choke.
I did use the Lundahl LL2733 before and there are some shops in Germany who offer different current rating ( so less current by reducing the airgap will give you more mH ). DCR will stay the same but getting more mH with the same mH rating will be benificial.
The lundahls have two coils that can be put in series or in parallel. The LL2733 in parallel is 0,85 ohm . I think that is ok. In series it will be 2*1,7 ohms so 3,4 ohms which seems to high to because of the flucuating current being drawn.
I think 0,85 ohm is low enough to keep using the present transformer . Maybe replace the original diodes by schottky.
The original transformer is rated 58,5 volt 3,1A with a centre tap.
The shop in Germany tells us on their website that the Lundahls can take more current when the ac voltage across the choke is lower. The Ll2733 is specified for 60 volts rms so in a CLC there will be a lot less. The LL2733 when in parallel is 100mH 3,4 A with 10 volt AC it can handle 5,4 A according to their website.
NOW the questions:
What current rating would be the minimum i will need for this amp in a CLC set up? So i can go for the highest mH and still be safe.
Would a choke input be usefull? So far i only used them for devices where the current drawn is pretty steady. With a low number of mH you will need to use a bleeder to draw a rather high current to make it work as a choke input and using the two coils of a LL2733 in series to get more mH will give you 3,4 ohms which will give you a BIG voltage drop. ( using LL2733 in my dddac that draws about 1A so i am aware of that)
Maybe use CLC and a Belleson regulator. There have been people using that one in a big Hafler amp.
Thanks for your input, Eduard
I just received a Mcintosh MC2505 stereo 50 watts power amp.
It has a seperate winding to supply the output stage with + and - 40 volts DC.
Power consumption is 250 watts at rated output so it will be class AB or just B?
Because i did use chokes in the past for my 30 watts Hiraga amp i wanna try adding two chokes to change the present power supply from just a single cap per rail to a CLC set up.
The original cap is 10000 µF so i will probably just add another and a choke.
I did use the Lundahl LL2733 before and there are some shops in Germany who offer different current rating ( so less current by reducing the airgap will give you more mH ). DCR will stay the same but getting more mH with the same mH rating will be benificial.
The lundahls have two coils that can be put in series or in parallel. The LL2733 in parallel is 0,85 ohm . I think that is ok. In series it will be 2*1,7 ohms so 3,4 ohms which seems to high to because of the flucuating current being drawn.
I think 0,85 ohm is low enough to keep using the present transformer . Maybe replace the original diodes by schottky.
The original transformer is rated 58,5 volt 3,1A with a centre tap.
The shop in Germany tells us on their website that the Lundahls can take more current when the ac voltage across the choke is lower. The Ll2733 is specified for 60 volts rms so in a CLC there will be a lot less. The LL2733 when in parallel is 100mH 3,4 A with 10 volt AC it can handle 5,4 A according to their website.
NOW the questions:
What current rating would be the minimum i will need for this amp in a CLC set up? So i can go for the highest mH and still be safe.
Would a choke input be usefull? So far i only used them for devices where the current drawn is pretty steady. With a low number of mH you will need to use a bleeder to draw a rather high current to make it work as a choke input and using the two coils of a LL2733 in series to get more mH will give you 3,4 ohms which will give you a BIG voltage drop. ( using LL2733 in my dddac that draws about 1A so i am aware of that)
Maybe use CLC and a Belleson regulator. There have been people using that one in a big Hafler amp.
Thanks for your input, Eduard
The minimum current rating for your choke should the maximum current your amplifier will draw, in.. say an overloaded amplifier. Add some 20% on top to be calm.
If you want to go for a choke input, the filtered output voltage will be nearly equal to 0.9*Vacrms instead of 1.41*Vacrms. So in order to keep the same rail voltage, your power transformer secondary must output 1.57 times higher voltage.
To ensure the choke input passive voltage regulation, the simple formula for the needed critical inductance is:
Lcrit = Vdc / Ima
If your rail DC voltage is 20V and you have a 100mH choke, then Icrit=200mA. Nothing fancy, the quiescent current of a class AB amplifier might do. I like to add 30% on top to not worry about it, so 260mA. Or a "bigger" choke.
Sonically wise, I favor choke input power supplies and I use them everywhere in my audio gear. They are quieter, they don't have the super-high order harmonics from the diodes switching, they have that passive regulation and subjectively, they sound much better to my ears. So my advice is to go for it.
I also recommend you to simulate your power supply using PSUD2. It's free to download and very simple to use. There you'll also check further details, such as the Q factor of the LC filter and the ripple.
About the decreasing of an air gap. This will increase the choke's inductance, but the maximum allowable DC current will be lower.
If you want to go for a choke input, the filtered output voltage will be nearly equal to 0.9*Vacrms instead of 1.41*Vacrms. So in order to keep the same rail voltage, your power transformer secondary must output 1.57 times higher voltage.
To ensure the choke input passive voltage regulation, the simple formula for the needed critical inductance is:
Lcrit = Vdc / Ima
If your rail DC voltage is 20V and you have a 100mH choke, then Icrit=200mA. Nothing fancy, the quiescent current of a class AB amplifier might do. I like to add 30% on top to not worry about it, so 260mA. Or a "bigger" choke.
Sonically wise, I favor choke input power supplies and I use them everywhere in my audio gear. They are quieter, they don't have the super-high order harmonics from the diodes switching, they have that passive regulation and subjectively, they sound much better to my ears. So my advice is to go for it.
I also recommend you to simulate your power supply using PSUD2. It's free to download and very simple to use. There you'll also check further details, such as the Q factor of the LC filter and the ripple.
About the decreasing of an air gap. This will increase the choke's inductance, but the maximum allowable DC current will be lower.
Hello,
You dont tell me something new because i already did use choke input in other things that i have build but never in something like this Mcintosh where i want to know the quiescent current and the maximum current.
So far i always did use a resistor to make sure that there is always the minimum current running.
Hope someone call tell me which current raiting to choose.
Greetings, Eduard
p.s of course i wanna try choke input too but then i will have to put an extra transformer. Maybe i will order an R core that will have several voltages and the secundairy side so i can switch to the right one
You dont tell me something new because i already did use choke input in other things that i have build but never in something like this Mcintosh where i want to know the quiescent current and the maximum current.
So far i always did use a resistor to make sure that there is always the minimum current running.
Hope someone call tell me which current raiting to choose.
Greetings, Eduard
p.s of course i wanna try choke input too but then i will have to put an extra transformer. Maybe i will order an R core that will have several voltages and the secundairy side so i can switch to the right one
simulate your proposals using PSUD2
you can compare output voltage and output ripple with different loads.
The last C in a multiple C PSU will supply most of the pulse current demanded by the amplifier.
Make sure it is big enough to maintain a nearly constant supply voltage at the amplifier.
In my opinion what you don't want is the amplifier's supply voltage to be determined by the (very variable) voltage drops of the preceding source impedances as set by the instantaneous current demands.
you can compare output voltage and output ripple with different loads.
The last C in a multiple C PSU will supply most of the pulse current demanded by the amplifier.
Make sure it is big enough to maintain a nearly constant supply voltage at the amplifier.
In my opinion what you don't want is the amplifier's supply voltage to be determined by the (very variable) voltage drops of the preceding source impedances as set by the instantaneous current demands.
Last edited:
Hello Andrew and all the others,
The French audiophiles did use 200mH chokes in a solid state 50 watt class A design some decades ago. They did use a regulated power supply for the input stage.
The Mcintosh however is no class A
The French design had a power supply with 32 volts dc output and 1,77A current drawn by the output circuit. Did use 33000µF then a 200mH and finally a pair of 100000µF.
Caps did get smaller but i will not be able to use caps like this in the present chassis.
Now there is just a 10000µF cap so making that one a bit bigger will not be a problem. The current will be provided by the last cap but i still wonder which current rating to use for the choke. If there will be to much current through the choke it will saturate and that will probably have side effects that we dont want.
The dcr of the choke the French did use was 0,7 ohm so i guess i will still be alright with the lundahl in parallel giving me 0,85 ohm.
The amp will be used for a horn system so the amp will have an easy life.
Maybe will try Belleson regulated power supply later on.
The French wrote in 1988 that using a regulated power supply for their 50 watt output stage was difficult. However they did use regulated supply for the input stage.
Maybe nowadays technology has evolved and things can be made that could notbe made some decades ago?
In the end the MC2505 will be used with a 12 db 600 hertz/cycles filter at its input so its task will not be that difficult i presume?????
Greetings, Eduard
The French audiophiles did use 200mH chokes in a solid state 50 watt class A design some decades ago. They did use a regulated power supply for the input stage.
The Mcintosh however is no class A
The French design had a power supply with 32 volts dc output and 1,77A current drawn by the output circuit. Did use 33000µF then a 200mH and finally a pair of 100000µF.
Caps did get smaller but i will not be able to use caps like this in the present chassis.
Now there is just a 10000µF cap so making that one a bit bigger will not be a problem. The current will be provided by the last cap but i still wonder which current rating to use for the choke. If there will be to much current through the choke it will saturate and that will probably have side effects that we dont want.
The dcr of the choke the French did use was 0,7 ohm so i guess i will still be alright with the lundahl in parallel giving me 0,85 ohm.
The amp will be used for a horn system so the amp will have an easy life.
Maybe will try Belleson regulated power supply later on.
The French wrote in 1988 that using a regulated power supply for their 50 watt output stage was difficult. However they did use regulated supply for the input stage.
Maybe nowadays technology has evolved and things can be made that could notbe made some decades ago?
In the end the MC2505 will be used with a 12 db 600 hertz/cycles filter at its input so its task will not be that difficult i presume?????
Greetings, Eduard
Hello Andrew,
BUT i dont know how much current there will be running.
At full power the amp takes 250 watts and the 58,5 volts with Centre tap has a 3,1 A current rating so i expect 3,1A will be the maximum current running through the choke because the transformer cannot deliver more.
Greetings, Eduard
BUT i dont know how much current there will be running.
At full power the amp takes 250 watts and the 58,5 volts with Centre tap has a 3,1 A current rating so i expect 3,1A will be the maximum current running through the choke because the transformer cannot deliver more.
Greetings, Eduard
Hello,
I have a 1,7 and a 1 A rated LL2733 so i can do a real set up in a single rail power supply with a big power resistor to make a big load like i did for my dac which uses some shunts.
It did show that when using the two coils in series voltage drop is big so with a fluctuating load like the mcintosh will be it will be better to use them in parallel.
Greetings, Eduard
I have a 1,7 and a 1 A rated LL2733 so i can do a real set up in a single rail power supply with a big power resistor to make a big load like i did for my dac which uses some shunts.
It did show that when using the two coils in series voltage drop is big so with a fluctuating load like the mcintosh will be it will be better to use them in parallel.
Greetings, Eduard
You do need to know the current flowing before you can make any use of PSUD.
If you do not already have a CRC filter in the power supply - make one and measure the voltage across the resistor. That will tell you what you need to know to get started.
You can work backwards by measuring the DC volts disconnected from the circuit and connected. Put the known components into the model and play with different resistor loads while taking note of the currrent going through the last/load resistor in the info panel of PSUD. You will find a load that corresponds to your DC voltage under load. You will then have a decent idea of the current flowing from your supply.
Having just completed phase one of my Mono J2 Project using the LUNDAHL chokes instead of a resistor. I am using the coils in parallel. Was worried they might be slightly stressed but they are staying cool. Your amp, being more powerful, will likely warm them up a bit more. AS the shop has said the LUNDAHLs are conservatively rated.
The coils in series will ot work well with a power amp THOUGH yours with the limited range you are using it in might not be as critical but nonetheless you will lose voltage and not likely gain much "goodness" for your trouble.
I had all kinds of kooky plans to get the ripply in the low nanovolts (as shown on PSUD) but in the end I realized I should use as few parts as possible and went with the tried and true CLC. I did use a big chunk of a can capacitor as the last C but my amp is used for 100 to 500 Hz and I figured that would work well in my case. I am assuming since you did not specify whether this was a low or high pass filter at 600Hz that it is a high pass. In that case, just a hunch, that maybe multiple paralleled smaller caps would be better for your application than a big, ol' can!
Be careful with your diode selection and consider using Mark Johnson's QUASIMODO to make a GOOD transformer secondary snubber. The info is here at DIYAudio if you have not run into it.
Take care,
If you do not already have a CRC filter in the power supply - make one and measure the voltage across the resistor. That will tell you what you need to know to get started.
You can work backwards by measuring the DC volts disconnected from the circuit and connected. Put the known components into the model and play with different resistor loads while taking note of the currrent going through the last/load resistor in the info panel of PSUD. You will find a load that corresponds to your DC voltage under load. You will then have a decent idea of the current flowing from your supply.
Having just completed phase one of my Mono J2 Project using the LUNDAHL chokes instead of a resistor. I am using the coils in parallel. Was worried they might be slightly stressed but they are staying cool. Your amp, being more powerful, will likely warm them up a bit more. AS the shop has said the LUNDAHLs are conservatively rated.
The coils in series will ot work well with a power amp THOUGH yours with the limited range you are using it in might not be as critical but nonetheless you will lose voltage and not likely gain much "goodness" for your trouble.
I had all kinds of kooky plans to get the ripply in the low nanovolts (as shown on PSUD) but in the end I realized I should use as few parts as possible and went with the tried and true CLC. I did use a big chunk of a can capacitor as the last C but my amp is used for 100 to 500 Hz and I figured that would work well in my case. I am assuming since you did not specify whether this was a low or high pass filter at 600Hz that it is a high pass. In that case, just a hunch, that maybe multiple paralleled smaller caps would be better for your application than a big, ol' can!
Be careful with your diode selection and consider using Mark Johnson's QUASIMODO to make a GOOD transformer secondary snubber. The info is here at DIYAudio if you have not run into it.
Take care,
Hello Rick,
I know from the start that i had to know the current flowing in the amp to use psud software so that why i did ask for someone to kind of calculate it with the info given so far.
I know installing a resistor in between two caps will allow me to meaure the current. BUT because this is no class A design the current will fluctuate a lot and that is why i like to know the maximum current so the chokes will not saturate.
AND i wanna use choke input because i think that is the best way to use a choke in a device with a steadsy current being drawn. However i dont know if it will also work for my amp .
The amp will be used up to 600 hertz/cycles . Choke input so far did always give me the impression that there is a lot more energy available, never running out of steam.
Will probably have to add another transformer ( probably 500 va Rcore) to use instead of the winding (58,5 volts 3,1A) . I can do some test with some transformers in my attick using the Lundahl chokes. I know that with choke input voltage drop can be high. I am kind of worried that voltage drop during high current peaks would make the operating voltage of the output stage to low but that couldnt be measured because my meter will be to slow to measure it?
The ll2733 in parall;el is only 0,85 ohm so difference in voltage drop between '' idle '' and heavy work would not be to big.
Thanks so far.
Greetings, Eduard
I know from the start that i had to know the current flowing in the amp to use psud software so that why i did ask for someone to kind of calculate it with the info given so far.
I know installing a resistor in between two caps will allow me to meaure the current. BUT because this is no class A design the current will fluctuate a lot and that is why i like to know the maximum current so the chokes will not saturate.
AND i wanna use choke input because i think that is the best way to use a choke in a device with a steadsy current being drawn. However i dont know if it will also work for my amp .
The amp will be used up to 600 hertz/cycles . Choke input so far did always give me the impression that there is a lot more energy available, never running out of steam.
Will probably have to add another transformer ( probably 500 va Rcore) to use instead of the winding (58,5 volts 3,1A) . I can do some test with some transformers in my attick using the Lundahl chokes. I know that with choke input voltage drop can be high. I am kind of worried that voltage drop during high current peaks would make the operating voltage of the output stage to low but that couldnt be measured because my meter will be to slow to measure it?
The ll2733 in parall;el is only 0,85 ohm so difference in voltage drop between '' idle '' and heavy work would not be to big.
Thanks so far.
Greetings, Eduard
Your maximum current will depend on how loud you listen and the impedance of your loudspeaker.
Incidentally what you said earlier -
i expect 3,1A will be the maximum current running through the choke because the transformer cannot deliver more
is mistaken. The transformer current rating is based on temperature rise - short-term your transformer can easily exceed the 3.1A rating.
If I were you I'd not be considering choke input at all for an amp which isn't classA. Use multiple stage LC with inductors which saturate relatively softly. I would suggest Sendust toroidal cores for best-bang-for-the-buck. If you have more money and hate winding then go for hi-flux cores.
Incidentally what you said earlier -
i expect 3,1A will be the maximum current running through the choke because the transformer cannot deliver more
is mistaken. The transformer current rating is based on temperature rise - short-term your transformer can easily exceed the 3.1A rating.
If I were you I'd not be considering choke input at all for an amp which isn't classA. Use multiple stage LC with inductors which saturate relatively softly. I would suggest Sendust toroidal cores for best-bang-for-the-buck. If you have more money and hate winding then go for hi-flux cores.
Hello,
Did find some info on this site:
http://www.diyaudio.com/forums/soli...ss-ab-power-amplifiers-rules-calculating.html
The MC2505 has an output transformer so maybe the current demanded from the power supply by the ooutput stage will be less fluctuating because of impedance irregularities of the loudspeaker.
My speaker has a nominal impedance of 16 ohm and the irregularities are not dips but peaks.
Maybe i will go with CLC and add a Belleson regulator. Rick who did also reply to my question did mention this oner in another thread. Maybe replacing the standard diodes with schottky diodes will give me enough voltage to use CLC and end up with the same output voltage.
Adding the Belleson will probably need another transformer because the drop across the regulator will be between 2 and 3 volts depending on the current being drawn.
I think you need to use air gapped chokes like the Lundahls to get some mH in order to create a useful filter.
greetings, Eduard
Did find some info on this site:
http://www.diyaudio.com/forums/soli...ss-ab-power-amplifiers-rules-calculating.html
The MC2505 has an output transformer so maybe the current demanded from the power supply by the ooutput stage will be less fluctuating because of impedance irregularities of the loudspeaker.
My speaker has a nominal impedance of 16 ohm and the irregularities are not dips but peaks.
Maybe i will go with CLC and add a Belleson regulator. Rick who did also reply to my question did mention this oner in another thread. Maybe replacing the standard diodes with schottky diodes will give me enough voltage to use CLC and end up with the same output voltage.
Adding the Belleson will probably need another transformer because the drop across the regulator will be between 2 and 3 volts depending on the current being drawn.
I think you need to use air gapped chokes like the Lundahls to get some mH in order to create a useful filter.
greetings, Eduard
Changing diodes
Hello,
I did open the MC2505 because i need to put the two primary windings in series to make it work at the 230 AC voltage in my country.
I can put two LL2733 on top of the chassis where the original caps are located now and put the new caps on the inside. Because i am in the '' metal construction '' i can make a new bottom plate with maybe some extra holes to allow the mounting of the caps. And the bottom plate will get a kind of hat to create an extra space for the caps and voltage dropping resistors for the input circuit power supply.
BUT first will power it up and connect before making any changes.
The 4 big diodes are still the original? ones from Motorola . It looks like M8207R product code. The transformer is 58,5 volt with a center tap. I wanna replace them with schottky to get some extra voltage to compensate for the use of the chokes ( 0,85 ohm DCR). Will it be enough? i still think that 39 volt witch LCL will be better than just a single cap with 40 volts.
Maybe will add a Belleson regulator later but then need more than 58,5 volts.
Greetings, Eduard
Hello,
I did open the MC2505 because i need to put the two primary windings in series to make it work at the 230 AC voltage in my country.
I can put two LL2733 on top of the chassis where the original caps are located now and put the new caps on the inside. Because i am in the '' metal construction '' i can make a new bottom plate with maybe some extra holes to allow the mounting of the caps. And the bottom plate will get a kind of hat to create an extra space for the caps and voltage dropping resistors for the input circuit power supply.
BUT first will power it up and connect before making any changes.
The 4 big diodes are still the original? ones from Motorola . It looks like M8207R product code. The transformer is 58,5 volt with a center tap. I wanna replace them with schottky to get some extra voltage to compensate for the use of the chokes ( 0,85 ohm DCR). Will it be enough? i still think that 39 volt witch LCL will be better than just a single cap with 40 volts.
Maybe will add a Belleson regulator later but then need more than 58,5 volts.
Greetings, Eduard
Attachments
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.